The Market for

Standard-Essential Patents

Justus A. Baron Laurie Ciaramella

Northwestern University Mines ParisTech

Transfers of Standard-Essential Patents

• Large transactions have attracted significant public attention

• Sale of Nortel portfolio of SEPs to a consortium including Apple, Ericsson,

Microsoft and Sony for 4.5 bn USD

• Purchase by Google of Motorola for 12.5 bn USD, resold (without patent

portfolio) for 2.91 bn USD two years later

• Policy concerns regarding licensing of SEPs on fair,

reasonable and non-discriminatory (FRAND) terms

• Antitrust investigation of SEP transfers raising the prospect of circumventing the

original owner’s licensing obligations (Bosch to IPcom, National Semiconductor

to N-Data)

• Significant proportion of SEP litigation initiated by patent assertion entities and

other entities who purchased already established SEPs (Contreras, 2016)

\hline

Justus A. Baron

The Market for Patents

• Facilitating the transfer of technologies is a core function of the patent

system (Spulber, 2015)

• The re-sale market for patents is a part of the market for technologies

• Significant potential for welfare gains from re-allocation of ideas (Serrano, 2011; Akcigit

and Kerr, 2015)

• Market for technologies comprises licensing, cross-licensing, transfer of ownership

(Arora et al., 2004; Gambardella et al., 2007, Arque-Castells and Spulber, 2017)

• Transfer of knowledge from inventor to innovator (Figueroa et al., 2013)

• The market for patents is also a market for the right to assert (Galasso

et al., 2013, Gaessler, 2016)

\hline

Justus A. Baron

Specificities of SEPs

• Standardization partly overcomes potential for misallocation of ideas

• Technology standards should be fully open to implementers, and provide all necessary

technical information

• SEPs are generally subject to FRAND licensing obligations

• Can’t be used to exclude rivals from using the technology

• Obligations “travel” with the SEP in case of transfer

• Standards are often subject to multiple SEPs owned by different firms

• SEPs don’t confer an exclusive right over the standardized technology

• Freedom to operate and assertion efficiencies as motives for SEP transfers?

\hline

Justus A. Baron

Possible reasons for SEP transfers

• Enforcement: Majority of SEP infringement litigation brought by

assertion specialists (Contreras, 2016)

• Freedom to operate: Defensive acquisitions by implementers and

aggregators (Cosandier et al., 2014)

• Aggregation: vertical integration is the textbook solution to royalty

stacking and transaction costs resulting from fragmentation

• Privateering: e.g. Ericsson to Unwired Planet; attenuate FRAND

limitations and potential repercussions on business relationships from

aggressive enforcement

• Vertical specialization: a limited number of firms (“standardization

insiders”) account for majority of contributions to SSOs

• May acquire patents for introduction into standards, and sell once essential

\hline

Justus A. Baron

Empirical contributions of our paper

• We compare assignee and assignor characteristics:

• SEP Portfolio size: we test whether re-assignments reduce or increase concentration

of SEP ownership

• SSO membership & contributions: Standardization “insiders” and “outsiders”

• Standard-compliant products: implementers vs. non-practicing entities

• We compare characteristics of firms participating in the “ex ante” and

“ex post” market

• Ex ante: transfers of patents before declaration as SEP

• Ex post: transfers of declared SEPs

\hline

Justus A. Baron

Empirical methodology

• We use data on SEP declarations from Baron and Pohlmann (2017)

• SSO policies require or encourage declaration, based on personal knowledge, no 3rd

party evaluation

• 9,155 unique US patents declared essential to various SSOs

• We use USPTO reassignment data to study patent transfers

• Widely used in the literature (Serrano, 2010, 2011; Galasso et al., 2013; Figueroa et

al., 2014; Akcigit and Kerr, 2015; Ciaramella, 2017; Arque-Castells and Spulber, 2017)

• 1,629 SEPs involved in 2,580 transfers (excluding intra-firm and multiple transfers in

single year)

• SSO contributions and membership from Searle Center Database

(Baron and Gupta, 2017; Baron and Spulber, 2017)

\hline

Justus A. Baron

Declared SEPs in sample, by SSO

\hline

Justus A. Baron

Evolution of Firms’ SEP Portfolio

• Do patent reassignments contribute to increase (aggregation) or reduce

(privateering) concentration of SEP portfolios?

• We define a firm’s SEP portfolio as

• 𝑆𝑖,𝑡 = 𝑆𝑖,𝑡−1 + 𝐷𝑖,𝑡 − 𝐸𝑖,𝑡 + 𝑃𝑖,𝑡 − 𝑉𝑖,𝑡

where D is the number of declarations, E the number of elapsed and expired SEPs, and

P and V respectively the number of SEPs acquired and sold

\hline

Justus A. Baron

Justus A. Baron

Justus A. Baron

Justus A. Baron

Justus A. Baron

Justus A. Baron

Justus A. Baron

Evolution of Firms’ SEP Portfolio

• Do patent reassignments contribute to increase (aggregation) or reduce

(privateering) concentration of SEP portfolios?

• We define a firm’s SEP portfolio as

• 𝑆𝑖,𝑡 = 𝑆𝑖,𝑡−1 + 𝐷𝑖,𝑡 − 𝐸𝑖,𝑡 + 𝑃𝑖,𝑡 − 𝑉𝑖,𝑡

where D is the number of declarations, E the number of elapsed and expired SEPs, and

P and V respectively the number of SEPs acquired and sold

• The SEPs enter the portfolio of the assignee and exit the portfolio of the

assignor with the transfer; we thus compare the portfolio size of the

assignor before with the size of the assignee’s portfolio after the transfer

\hline

Justus A. Baron

Assignor and assignee portfolio size

Justus A. Baron

Timing of assignment wrt. declaratoin

Justus A. Baron

Insiders & outsiders, before & after

declaration

Justus A. Baron

Insiders & outsiders, before & after

declaration

Justus A. Baron

Econometric implementation

• We create a sample of patent-SSO-firm observations:

• transferred patent i can be declared to multiple SSOs s; and we observe each transferred

patents-SSO observation for both the assignee and the assignor

• We estimate the following regression equation

𝑌𝑖,𝑠,𝑡 = 𝛽1𝐴𝑠𝑠𝑖𝑔𝑛𝑒𝑒𝑖,𝑗,𝑡 + 𝛽2𝐵𝑒𝑓𝑜𝑟𝑒𝑗,𝑠,𝑡 + 𝛽3𝐴𝑠𝑠𝑖𝑔𝑛𝑒𝑒 𝑥 𝐵𝑒𝑓𝑜𝑟𝑒𝑖,𝑗,𝑠,𝑡

+𝛽4𝑍𝑖,𝑗,𝑠,𝑡 + 𝛽5𝑋𝑡 + 𝛽6𝑊𝑠 + 𝜀𝑖,𝑗,𝑠,𝑡

where Y is a vector of explained firm characteristics, Z is a vector of control variables, X

and W respectively are vectors of year and SSO fixed effects.

\hline

Justus A. Baron

Justus A. Baron Producer SDO

member

# member

ships

# contri-

butions

SEP

portfolio

Assignee 0.020

(0.05)

-0.084*

(0.05)

-8.143*

(4.03)

-1500***

(487)

11.88

(11.6)

Before -0.096**

(0.04)

0.015

(0.05)

-4.080

(2.58)

-543**

(267)

12.28

(13.0)

Assignee x Before 0.022

(0.05)

0.056

(0.07)

5.178

(4.65)

1483**

(481)

-5.16

(21.6)

Producer 0.778***

(0.03)

32.016***

(3.44)

2476***

(888)

27.60

(25.6)

SDO member 32.706***

(3.17)

-529

(453)

90.88***

(29.7)

# memberships 58**

(11)

-0.87***

(0.2)

# contributions 0.01***

(0.0)

Grant lag -0.000

(0.00)

0.000

(0.00)

-0.001

(0.00)

0.353***

(0.11)

0.002

(0.00)

Constant -0.036

(0.11)

-0.028

(0.07)

8.853

(8.97)

-1926*

(1067)

-18.963

(31.16)

N 4,796 4,796 4,796 4,796 4,796

r2 0.266 0.613 0.633 0.474 0.504

* p < 0.1, ** p < 0.05, *** p < 0.01

Conclusion

• Patents are transferred from standardization “insiders” to “outsiders”

after becoming standard-essential

• This pattern is not confirmed for patents transferred before essentiality declaration

• Standard implementers participate more significantly in the market for

already declared SEPs, but on both sides of the market

• SEP transfers neither reduce nor increase the extent of concentration

of SEP ownership

• No evidence for either privateering or aggregation; but rather vertical

specialization from invention to standardization and beyond

\hline

Justus A. Baron